Is traditional ECM up for high performance 9 March 2009
Posted by lopataru in ECM.Tags: ECM, performance
2 comments
Last week I stumbled over the WordPress statistics for February.
Impressive. That got me thinking on how I would be able to implement this kind of backend system with a traditional (read “top Gartner stuff”). And I shivered inside while thinking of IBM CM, Documentum, FileNet… SharePoint (lol!).
I remember once I’ve seen a support issue with one of the above vendors in which the administration tool could not display the size of filestores bigger that 2 GB. And I believe the issue is still there, after 2-3 years. Tragic.
Imagine a customer seeing this (and they do) and saying… “Well.. what kind of Enterprise system is this? If it cannot show correctly storage spaces over 2 GB? How can I trust it with my Terabytes?”
My practice based experience tells me that a large scale performance cannot be normally achieved with Enterprise grade software. You have a better chance with some high skilled professionals (not many, 4-7 should be enough) which can put together some “indie” software.
In my work, I rarely (read “never”) seen a ECM system handle a load similar with the WordPress one. ECM in my area tends to cap normally at about several million items and a few TB of space. While requiring a huge deluge of hardware (I can’t still understand why on earth would I need minimmum 4 GB or RAM for an Index Server? Even in the smallest install)
You out there…. Working on ECM… I would like to know your statistics.
Documentum 6.5 ramblings, or something 2 March 2009
Posted by lopataru in ECM.Tags: documentum
add a comment
I just noticed that another month passed by without me writing anything here.
Obviously because I’m busy… same (lame) excuse. Not really. Of course the job is always demanding, but getting 30 minutes a day to ramble about something can’t be that hard.
So I pulled myself and logged back here. Let’s talk about my recent Documentum experiences.
At this stage we are undergoing some D 6.5 implementations. SP1, just to be sure.
I’m not involved technically first hand, and the last time I got my hands on it was about 4-5 months ago. This is to settle expectations straight… you’re not about to see any technical revelations here.
First thing: “shit! is different!” the whole install process I mean. Where’s my DocApp? No more docapps, use the “.dar”. Learn Eclipse… headless.. that is. The Eclipse, not me. Not yet.
Ok, basic stuff works. TaskSpace… almost ok. Let’s move to the Imaging Services… Brr… Strike 1… Strike 2… Use Webtop.
I promise myself I’ll get back on this. Never did. And now my colleagues suffer the same things. Poor them.
But it moves faster. Lot faster. Then I increased the RAM available to the virtual machine. Shouldn’t have done that.
Note to self: don’t increase the memory for virtual machines upon the exhaustion of the host memory. Swapping is bad..
Business Process Manager now. Done. All ok. BP Services? Not yet. Later.
Oooo… BAM. I mean Business Activity Monitor. I feel like Dee-Dee in the laboratory. Same effect.
Rollback to VM snapshot.
At this point I think: there must be an easy way.
Then I read in the installation manual: “rename the XXX file to YYY file and then run setup.exe” (names were changed to protect the innocent). As a last line on the page. Why the %^&*() do I need to rename a file which is provided in the installation kit? Oh well.. this why we are payed the good bucks.. can only imagine how SAP looks like.Oh, stop, I know that also… tough job.
Wanna try Annotation Services? With 6.5 ? Sure…. bring it on. It worked! That brought the spirit up. For a moment.
Today it does not work anymore. Looking for some LiveCycle piece? Good luck!
Great. Now what? Laugh hysterically and get back on it.
You know, Oracle installation kits are free on the Internet. You know why? Because it takes a skilled person to install it properly. Go figure.
Why am I telling you this? Because next week when all the puzzle pieces will be nicely put toghether I can look here with my colleagues and laugh.
And since you read it all the way up to here: let’s build a HA environment toghether. And put some Document Sciences on it to spice it up.
PhD paper done. Phew! 19 January 2009
Posted by lopataru in ECM, Research.3 comments
Finally, after a long time i have now a complete version of my PhD thesis.
I would like to ellaborate more on this, but after spending 4 days in a mountain cottage secluded in front of my laptop… i simply can’t.
Now i just need to publish some articles and present my creation to the public. Behold 
Kickoff 2009 3 January 2009
Posted by lopataru in ECM, Various.add a comment
Just entered 2009 and making new year resolutions? I hope I’m past that moment since i just read somewhere that such decisions are bad for your health.
What will 2009 mean for Content Management? Nothing special, I guess.. But i think the whole IT industry is going to pass through some good times. In the last few years i’ve seen some decay in the overall industry practices. A lot of marketing and little substance. Everybody wanted to be ahead of the competitor, to show (keyword: “show”) it’s the best… blah blah blah..
I expect 2009 to be an year in which the industry is forced to mature itself.. maybe because of the so much mediatized financial crisis . Thinking of it, this crisis reminds me of Share Point. lol
In the CM field we will probably see a lot of talk around “compliance” solutions.
A lot of talk around increased efficiency (everyone will want that, won’t they?).
The major ECM vendors (IBM, EMC, Oracle, MS) will try to grow their direct consulting practice to snatch a bigger part of the pie (not only licenses).
A lot of focus will go to open source (it’s “cheap”, isn’t it?).
Outsourcing will get more consolidated to major players. I see many small single-client outsourcing companies slowly dying. Which will bring more qualified personnel to the market. Please, do come in!
Cutting costs? Only the non-productive ones probably. Who is smart should invest in R&D at these times. And collect on its existing knowhow.
Low budgets from customets? Eh.. what’s new ? it was always this way. Maybe they say it a little more these days, but we should get used to smaller budgets until we can provide really good quality (nobody argues about price when buys a Bentley or Ferrari).
So, what’s in for 2009? Nothing special, just aim to be better.
Happy New Year!
Snail install 21 November 2008
Posted by lopataru in Various.4 comments
Today I have been involved in a product upgrade on the Documentum platform.
It was a service pack install, usually not very complicated. Although this blog is about content management… i’ll not talk about this right now. In fact, i will talk about the upgrade experience.
Before the installation i wrote down a complete procedure to follow. The installation itslef would be done by a person which is not very familiar with Documentum so it was pretty explanatory.
The installation needed these kind of operations:
- a lot of file and folder copying
- various users to be used
- GUI as well as command line operations
- file contents comparison
- file edits
- folder contents comparison
- restore from backup archives
And now i come to the pain point: these needed to be done remotely on Unix machines.
On my “home” Unix environment the operations lasted about 2 hours, without rush.
In the today’s installation they lasted 7 hours.
Let’s say that the person performing the upgrade procedure was not very talented although he was quite knowledgeable at Unix (quite fluent in command line tricks).
I think the reason is the fact the Unix environment (Solaris based) did not offer enough support for such operations. This meant that the operations which needed to be done became very labour intensive for a “regular” admin.
I give some examples:
- compare 2 folder contents and synchronize them using *nix native commands
- compare 2 files using only less/cat and vi
- move files between complex folder structures named in a similar way using command line
On my “home” environment i have some nice tools to make this easier, but on a corporate, “bare” environment these tools do not exist and you need to rely on the trustful vi/grep/cp/less/cat commands. While these commands are excellent and definitely more powerful than the normal Windows counterparts, the effort needed from an average user (let’s not call it “admin”) is far bigger.
Since all of these operations needed to be done like this, it needed extensive attention (looking at each command some tens of seconds and making checks and double-checks to make sure the result is indeed what we needed).
My conclusion is that even though Unix is arguably better than Windows on server side, you need a highly skilled personnel to operate it and to make use of proper aditional tools (or scripts) to administer it efficiently.
I’m sure any *nix expert out there would jump up and bash me for saying this, and would be ready to prove to me he can do anything faster in *nix that on Windoze. I agree with this. But I’m talking about the average dude operating the corporate infrastructure. Keywords: “average”, “dude” and “corporate”.
Am I really wrong? Or *nix installs of software usually take longer than Redmond ones?
Momentum 2008 – Closing notes 14 November 2008
Posted by lopataru in ECM.Tags: emc, events, momentum 2008
add a comment
It ended today.
Generally I think it was a very good event. About 1300 persons attended and it was a good mix of customers / partners / EMC.
As I sit in the hotel room right now I try to think if there are any highlights still alive in my mind… please don’t mind the somehow incoherent ramblings below… heck.. it’s just a blog.
One rather small dissapointment was the fact that CMIS did not get too much coverage. Except some mentions done especially by Andrew Chapman (and maybe Mark Lewis… I can’t remember) I’ve not noticed anything special. Pity.
On the upnote, i did not encounter the SharePoint word as much as in other events / presentations. I guess the hype has somehow settled and now everyone is minding their own business in the usual way.
Very nice evening events… One at the Castle, one at the Municipal House. I love Europe . And especially Prague as a matter of fact. No.. i don’t live in Prague…
Some very interesting presentations from Ed Bueche on the performance side of things. Batching… scoping and high volume server…
I’m not yet clear on what to do with all the SP-DCTM integration opportunities i have on my table. The EMC products on this side are promising.. but the roadmap… hmmm… The partner offering is really only from one company (from the exibitors list). The rest are just plain simple (as I would build it in a few weeks).
As I am on this topic, mind this: UCF .NET
I’ve seen also a lot of talk around Captiva Dispatcher. Nice, useful module for high volume paper processing. Still somethings missing there, but oh well.. you can’t have it all. Or can you?
If I would be to give a prize to the thing which raised my eyebrows the most: www.nnit.com/continuum. I just loved the eye-candy: it’s a very nicely done UI. Yes, i know… i’m shallow.
See you next year, after we manage to make a buck or two (euro in my case) in the current financial downturn.
Momentum 2008 – First days 12 November 2008
Posted by lopataru in ECM.Tags: documentum, ECM, momentum 2008
add a comment
These days I am in Prague and attending the EMC Momentum.
It’s the first time for me at Momentum although i have been attending the US World shows in the past years.
I planned to write a review after each day (today is the 3rd one) but the combination of Staropramen and customers filled up all my evenings.
When i go to such events i’m always traing to get a feeling of the “vibe” and vision coming from the speakers. I’m usually searching for the hints to define the long term stategy. In the recent past i attended the IBM show and obviously i’m comparing this one to it.
I don’t know if it’s the financial crisis or something in the air… but the vibe was not here. Don’t get me wrong, the vision is fine. All the guys & galz presenting were saying very sensible things and in a professional way. I just missed the enthuziasm.
What’s the vision? The usual: go from data to information and knowledge. I don’t particulary like the name “Knowledge Worker” but if this is the new marketing buzz.. then why not.
The highlights:
- high volume server – really good for high volume ingestion. Yes, everyone talked about ingestion only. What to do after you get it in? Nobody knows. i think this should be addressed to. I can’t imagine doing a faceted search on billion of objects . Hm… let’s put CenterStage on top of HA Server ….
- XML Store – this is a big thing. I don’t want to act as a future teller (although we’ll have a futurologist here at Momentum) but i think this will be the next default storage space of Documentum. I’ve shared this thought with Michiel Veen in the second day and it seems is not just an idea. I’ll keep an eye on it.
- SaaS – Not just a hype, in the keynote this was addressed in a very good way from my point of view. I plan a sepparate post on this
- CenterStage – definitely the star of the show. This is the new client for Documentum. Webtop / WDK lived its life and now it’s the time to give way to the more Apple-like UI. Obviously this will take tremeduous effort from the community since there are a lot of solution built on WDK which will need to be redone in an entirely new web way. I think this is good – i’m not a very fond person on the “backward compatibility” idea in UI.
- Victor Spivak – the architect of the house. As usual, the best technical presentation in the house. It was a shame that the presentation was stopped before finish. The whole packed room would have wanted to see more (and Victor was eager to tell it all) but the lady in the back keep insisted to close the session (dunno way, just to make the scheduled time i guess)
- Elephant in the Room – i was in the SP Advisory Forum (hi Robin!) and got into some interesting discussions on the integration topic. Robin has it covered in its post and i think it makes a very good description of the session. I was pretty vocal on the need to really reuse the SP content once it arrives in Documentum since i really believe customers don’t care what system “owns” the content and they just want to work with it from the application of choice. More on this in another post.
now i’m on my way to the Sponsor Track. See ya!
Practical scalability in CMS 31 October 2008
Posted by lopataru in ECM.Tags: scalability, sharepoint
add a comment
Couldn’t resist in posting this…
One colleague came to me today and described a problem (and a resolution):
SharePoint. over 6 million documents inside. a “few” GB in size. what to do?
The problem was, the client needed to move this content volume from one installation to another. The dump&load tools which is from Microsoft seems to take each content item and shove it to the dump. With probably a lot of overhead in each get next, fetch, load, save, close… Sometimes crashing… eh.. that’s life
What to do? Get the content out of SharePoint ! Now, please. (awating flames on this…)
Round 2:
SharePoint. Gimme a “count” of all items in a document library… “Out of Memory”. whtf?!
Why? The CMS (lol) fetches each item and “counts” it. Like using your fingers to count.
What to do? Beats me.. Ask some moooore money to improve it through development? Lol
Rant done. Flame away!
Yet Another CM standard 14 September 2008
Posted by lopataru in ECM.Tags: cmis, ECM
add a comment
CMIS was submitted to Oasis for becoming an open standard on Content Management Interoperability.
What’s fun, is that almost every single blog / newsfeed I monitor in the Content Management area addressed the subject. It was fun to see my dashboard with the acronym popping up in all sections (i keep it organized by different vendors).
This definitely got the attention of the ECM professionals. I’ll not write too much on it since a lot of bytes are already out there commenting the event.
What I want to point is that this needs the proper amount of luck and environment in order to succeed (unlike JSR 170 and 283).
It needs to be the SQL of Content Management. On one blog entry, Pie noted that the new standard needs support from the vendors in order to make it into the real world. This is essential of course, but another thing is to also be embraced by the application builders as a core CM layer of their architecture.
Declarative support is there (e.g. Microsoft) . At least from the vendors – the first being obviously the standard promoters. But even most of them were giving only the marketing speech (with the notable exception of Alfresco who already has an implementation).
I think this is a great opportunity for the ones building CEVAs as it might reduce their costs on supporting several platforms for their solution. Unfortunately there is a difference between SQL and CMIS – databases are well established in IT architectures as foundation technologies for various solutions. Content Management is more seen as a top level application.
It will take time for IT Managers, Solution Architects, CIOs, ISVs and others to look at CM and think of it as middleware (at least, if not infrastructure) in order to start building applications which are independent of the ECM product.
Clouds, virtualization and other hip stuff going kaboom 13 August 2008
Posted by lopataru in Various.Tags: cloud computing, crash, IT strategy
add a comment
After reading the news the other day, i had an idea to put a blog entry but never got to.
Today I saw a post on bmoc and I began writing a comment on it. Was too long, so i evolved it to a blog post:
The recent mediatized cloud crash of Linkup and also the massive VMWare bug comes to prove that “innovative” IT technologies come with an inherent risk of their own.
Both had somewhat straighforward solutions (some of the many):
1. just have a decent backup solution.
2. don’t use license management in enterprise class software.
But technology cannot solve all on its own. You need other enforcement which works in real life. Like Marko showed with the US Postal Service sample.
It so happens that the company I’m in provides also some sort of “cloud” services. I’m not going to say we’re better than others, but one thing i know we have is insurance and this is included in one to one contracts with all cloud customers (each contract is tailor made, not just an “accept terms and conditions” web check box).
Of course, in case of a disaster no one can provide the lost data (if it’s really lost meaning no backup exists).
Therefore, I think that anyone which does not complete it’s IT infrastructure with additional material procedures likes to live on the edge.
On a last minute info, i have just learned that one of our customers was running its servers on a RAID array which was since several months in “destroy” mode since all spare and redundant hdd’s were dead and no one noticed. Comes to prove the point.
If i were to be sadic, i would say “no problem, more money for us.. we’ll just charge a hefty sum to rebuild its infrastructure”. But the problem is it never should have happened in some persons would have acted more professional and think outside the technology for a minute.
